<div dir="ltr"><div><div>Hi Yohei,<br><br></div>I see the issue. <br><br></div>Internally, with the 'taylor_stats' function contains:<br><div><div><br></div><div>[snip]<br></div><div> else ; CURVILINEAR GRID: w[*][*]<br> tmean = wgt_areaave2(t, w, 1) ; area weighted means<br> rmean = wgt_areaave2(r, w, 1)<br></div><div>[snip]<br><br></div><div>The last argument of the 'wgt_areaave2' function ' 'opt' ] should be set to 0 not 1<br><br>From 'wgt_areaave2' documentation:<br><br><em>opt</em>
<p class="gmail-indent">If <em>opt</em> = 0, the area average is calculated using available non-missing data.
If <em>opt</em> = 1, then if <b>any</b> point in <em>q</em> is missing, the area
average is not computed. In this case, it will be set to the missing value, which is
indicated by <em>q</em>@<a href="http://www.ncl.ucar.edu/Document/Language/fillval.shtml">_FillValue</a>, or the default missing value if
<em>q</em>@_FillValue is not set.</p><p class="gmail-indent">=====</p><p class="gmail-indent">I made the simple change and have attached the code. You can load directly:</p><p class="gmail-indent">load "./taylor_stats.ncl"</p><p class="gmail-indent">[your code]<br></p><p class="gmail-indent">Please let ncl-talk know if this fixes the problem. <br></p>=====<br><br></div><div>Cheers<br></div><div>Dennis<br></div></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Jul 28, 2017 at 9:44 AM, Yohei Takano <span dir="ltr"><<a href="mailto:yohei.takano@mpimet.mpg.de" target="_blank">yohei.takano@mpimet.mpg.de</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear NCL users,<br>
<br>
I am trying to use "taylor_stats" function to calculate pattern<br>
correlation, ratio, and bias.<br>
<a href="https://www.ncl.ucar.edu/Document/Functions/Contributed/taylor_stats.shtml" rel="noreferrer" target="_blank">https://www.ncl.ucar.edu/<wbr>Document/Functions/<wbr>Contributed/taylor_stats.shtml</a><br>
<br>
The two ocean data (test array and reference array) are both<br>
curve-linear grid and contains missing values for land. Reading the<br>
descriptions in NCL document, I think I need to apply spatial weights to<br>
the data so I prepared two-dimensional area data as a weight based on<br>
curve-linear grid information.<br>
<br>
Then I tried<br>
<br>
$ tay_stats = taylor_stats(test,reference,<wbr>area,0)<br>
: again test, reference, and area are two dimensional curve-linear grid<br>
data)<br>
<br>
but I have an error:<br>
$ fatal:The result of the conditional expression yields a missing value.<br>
NCL can not determine branch, see ismissing function<br>
<br>
The function works when I try<br>
$ tay_stats = taylor_stats(test,reference,1.<wbr>0,0)<br>
(for no weighting)<br>
<br>
so I am pretty sure it is about the weighting problem.<br>
Are we not allowed to include missing values in "area" for taylor_stats<br>
option? Since this is an ocean data, the area also contains missing<br>
values for land so I was wondering how I can deal with this issue.<br>
<br>
I could not attach the sample data (since it was a bit too big for the<br>
mailing list so I canceled it) but I could share it in some other way if<br>
necessary.<br>
<br>
If you could give me any suggestions, point out what I am missing or<br>
mistaking that will be great. Thank you for your help in advance.<br>
<span class="HOEnZb"><font color="#888888"><br>
Yohei Takano<br>
<br>
--<br>
Yohei Takano Ph.D<br>
Max Planck Institute for Meteorology<br>
Bundesstraße 53, Room 207<br>
20146 Hamburg, Germany<br>
Tel: <a href="tel:%2B49%20%280%2940%2041173%20-%20150" value="+494041173150">+49 (0)40 41173 - 150</a><br>
Email: <a href="mailto:yohei.takano@mpimet.mpg.de">yohei.takano@mpimet.mpg.de</a><br>
______________________________<wbr>_________________<br>
ncl-talk mailing list<br>
<a href="mailto:ncl-talk@ucar.edu">ncl-talk@ucar.edu</a><br>
List instructions, subscriber options, unsubscribe:<br>
<a href="http://mailman.ucar.edu/mailman/listinfo/ncl-talk" rel="noreferrer" target="_blank">http://mailman.ucar.edu/<wbr>mailman/listinfo/ncl-talk</a><br>
</font></span></blockquote></div><br></div>